summaryrefslogtreecommitdiffstats
path: root/third_party/libwebrtc/rtc_base/virtual_socket_server.cc
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 19:33:14 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-07 19:33:14 +0000
commit36d22d82aa202bb199967e9512281e9a53db42c9 (patch)
tree105e8c98ddea1c1e4784a60a5a6410fa416be2de /third_party/libwebrtc/rtc_base/virtual_socket_server.cc
parentInitial commit. (diff)
downloadfirefox-esr-upstream.tar.xz
firefox-esr-upstream.zip
Adding upstream version 115.7.0esr.upstream/115.7.0esrupstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'third_party/libwebrtc/rtc_base/virtual_socket_server.cc')
-rw-r--r--third_party/libwebrtc/rtc_base/virtual_socket_server.cc1383
1 files changed, 1383 insertions, 0 deletions
diff --git a/third_party/libwebrtc/rtc_base/virtual_socket_server.cc b/third_party/libwebrtc/rtc_base/virtual_socket_server.cc
new file mode 100644
index 0000000000..efc206b219
--- /dev/null
+++ b/third_party/libwebrtc/rtc_base/virtual_socket_server.cc
@@ -0,0 +1,1383 @@
+/*
+ * Copyright 2004 The WebRTC Project Authors. All rights reserved.
+ *
+ * Use of this source code is governed by a BSD-style license
+ * that can be found in the LICENSE file in the root of the source
+ * tree. An additional intellectual property rights grant can be found
+ * in the file PATENTS. All contributing project authors may
+ * be found in the AUTHORS file in the root of the source tree.
+ */
+
+#include "rtc_base/virtual_socket_server.h"
+
+#include <errno.h>
+#include <math.h>
+
+#include <map>
+#include <memory>
+#include <vector>
+
+#include "absl/algorithm/container.h"
+#include "api/units/time_delta.h"
+#include "rtc_base/checks.h"
+#include "rtc_base/event.h"
+#include "rtc_base/fake_clock.h"
+#include "rtc_base/logging.h"
+#include "rtc_base/physical_socket_server.h"
+#include "rtc_base/socket_address_pair.h"
+#include "rtc_base/thread.h"
+#include "rtc_base/time_utils.h"
+
+namespace rtc {
+
+using ::webrtc::MutexLock;
+using ::webrtc::TaskQueueBase;
+using ::webrtc::TimeDelta;
+
+#if defined(WEBRTC_WIN)
+const in_addr kInitialNextIPv4 = {{{0x01, 0, 0, 0}}};
+#else
+// This value is entirely arbitrary, hence the lack of concern about endianness.
+const in_addr kInitialNextIPv4 = {0x01000000};
+#endif
+// Starts at ::2 so as to not cause confusion with ::1.
+const in6_addr kInitialNextIPv6 = {
+ {{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2}}};
+
+const uint16_t kFirstEphemeralPort = 49152;
+const uint16_t kLastEphemeralPort = 65535;
+const uint16_t kEphemeralPortCount =
+ kLastEphemeralPort - kFirstEphemeralPort + 1;
+const uint32_t kDefaultNetworkCapacity = 64 * 1024;
+const uint32_t kDefaultTcpBufferSize = 32 * 1024;
+
+const uint32_t UDP_HEADER_SIZE = 28; // IP + UDP headers
+const uint32_t TCP_HEADER_SIZE = 40; // IP + TCP headers
+const uint32_t TCP_MSS = 1400; // Maximum segment size
+
+// Note: The current algorithm doesn't work for sample sizes smaller than this.
+const int NUM_SAMPLES = 1000;
+
+// Packets are passed between sockets as messages. We copy the data just like
+// the kernel does.
+class Packet {
+ public:
+ Packet(const char* data, size_t size, const SocketAddress& from)
+ : size_(size), consumed_(0), from_(from) {
+ RTC_DCHECK(nullptr != data);
+ data_ = new char[size_];
+ memcpy(data_, data, size_);
+ }
+
+ ~Packet() { delete[] data_; }
+
+ const char* data() const { return data_ + consumed_; }
+ size_t size() const { return size_ - consumed_; }
+ const SocketAddress& from() const { return from_; }
+
+ // Remove the first size bytes from the data.
+ void Consume(size_t size) {
+ RTC_DCHECK(size + consumed_ < size_);
+ consumed_ += size;
+ }
+
+ private:
+ char* data_;
+ size_t size_, consumed_;
+ SocketAddress from_;
+};
+
+VirtualSocket::VirtualSocket(VirtualSocketServer* server, int family, int type)
+ : server_(server),
+ type_(type),
+ state_(CS_CLOSED),
+ error_(0),
+ network_size_(0),
+ recv_buffer_size_(0),
+ bound_(false),
+ was_any_(false) {
+ RTC_DCHECK((type_ == SOCK_DGRAM) || (type_ == SOCK_STREAM));
+ server->SignalReadyToSend.connect(this,
+ &VirtualSocket::OnSocketServerReadyToSend);
+}
+
+VirtualSocket::~VirtualSocket() {
+ Close();
+}
+
+SocketAddress VirtualSocket::GetLocalAddress() const {
+ return local_addr_;
+}
+
+SocketAddress VirtualSocket::GetRemoteAddress() const {
+ return remote_addr_;
+}
+
+void VirtualSocket::SetLocalAddress(const SocketAddress& addr) {
+ local_addr_ = addr;
+}
+
+int VirtualSocket::Bind(const SocketAddress& addr) {
+ if (!local_addr_.IsNil()) {
+ error_ = EINVAL;
+ return -1;
+ }
+ local_addr_ = server_->AssignBindAddress(addr);
+ int result = server_->Bind(this, local_addr_);
+ if (result != 0) {
+ local_addr_.Clear();
+ error_ = EADDRINUSE;
+ } else {
+ bound_ = true;
+ was_any_ = addr.IsAnyIP();
+ }
+ return result;
+}
+
+int VirtualSocket::Connect(const SocketAddress& addr) {
+ return InitiateConnect(addr, true);
+}
+
+VirtualSocket::SafetyBlock::SafetyBlock(VirtualSocket* socket)
+ : socket_(*socket) {}
+
+VirtualSocket::SafetyBlock::~SafetyBlock() {
+ // Ensure `SetNotAlive` was called and there is nothing left to cleanup.
+ RTC_DCHECK(!alive_);
+ RTC_DCHECK(posted_connects_.empty());
+ RTC_DCHECK(recv_buffer_.empty());
+ RTC_DCHECK(!listen_queue_.has_value());
+}
+
+void VirtualSocket::SafetyBlock::SetNotAlive() {
+ VirtualSocketServer* const server = socket_.server_;
+ const SocketAddress& local_addr = socket_.local_addr_;
+
+ MutexLock lock(&mutex_);
+ // Cancel pending sockets
+ if (listen_queue_.has_value()) {
+ for (const SocketAddress& remote_addr : *listen_queue_) {
+ server->Disconnect(remote_addr);
+ }
+ listen_queue_ = absl::nullopt;
+ }
+
+ // Cancel potential connects
+ for (const SocketAddress& remote_addr : posted_connects_) {
+ // Lookup remote side.
+ VirtualSocket* lookup_socket =
+ server->LookupConnection(local_addr, remote_addr);
+ if (lookup_socket) {
+ // Server socket, remote side is a socket retreived by accept. Accepted
+ // sockets are not bound so we will not find it by looking in the
+ // bindings table.
+ server->Disconnect(lookup_socket);
+ server->RemoveConnection(local_addr, remote_addr);
+ } else {
+ server->Disconnect(remote_addr);
+ }
+ }
+ posted_connects_.clear();
+
+ recv_buffer_.clear();
+
+ alive_ = false;
+}
+
+void VirtualSocket::SafetyBlock::PostSignalReadEvent() {
+ if (pending_read_signal_event_) {
+ // Avoid posting multiple times.
+ return;
+ }
+
+ pending_read_signal_event_ = true;
+ rtc::scoped_refptr<SafetyBlock> safety(this);
+ socket_.server_->msg_queue_->PostTask(
+ [safety = std::move(safety)] { safety->MaybeSignalReadEvent(); });
+}
+
+void VirtualSocket::SafetyBlock::MaybeSignalReadEvent() {
+ {
+ MutexLock lock(&mutex_);
+ pending_read_signal_event_ = false;
+ if (!alive_ || recv_buffer_.empty()) {
+ return;
+ }
+ }
+ socket_.SignalReadEvent(&socket_);
+}
+
+int VirtualSocket::Close() {
+ if (!local_addr_.IsNil() && bound_) {
+ // Remove from the binding table.
+ server_->Unbind(local_addr_, this);
+ bound_ = false;
+ }
+
+ // Disconnect stream sockets
+ if (state_ == CS_CONNECTED && type_ == SOCK_STREAM) {
+ server_->Disconnect(local_addr_, remote_addr_);
+ }
+
+ safety_->SetNotAlive();
+
+ state_ = CS_CLOSED;
+ local_addr_.Clear();
+ remote_addr_.Clear();
+ return 0;
+}
+
+int VirtualSocket::Send(const void* pv, size_t cb) {
+ if (CS_CONNECTED != state_) {
+ error_ = ENOTCONN;
+ return -1;
+ }
+ if (SOCK_DGRAM == type_) {
+ return SendUdp(pv, cb, remote_addr_);
+ } else {
+ return SendTcp(pv, cb);
+ }
+}
+
+int VirtualSocket::SendTo(const void* pv,
+ size_t cb,
+ const SocketAddress& addr) {
+ if (SOCK_DGRAM == type_) {
+ return SendUdp(pv, cb, addr);
+ } else {
+ if (CS_CONNECTED != state_) {
+ error_ = ENOTCONN;
+ return -1;
+ }
+ return SendTcp(pv, cb);
+ }
+}
+
+int VirtualSocket::Recv(void* pv, size_t cb, int64_t* timestamp) {
+ SocketAddress addr;
+ return RecvFrom(pv, cb, &addr, timestamp);
+}
+
+int VirtualSocket::RecvFrom(void* pv,
+ size_t cb,
+ SocketAddress* paddr,
+ int64_t* timestamp) {
+ if (timestamp) {
+ *timestamp = -1;
+ }
+
+ int data_read = safety_->RecvFrom(pv, cb, *paddr);
+ if (data_read < 0) {
+ error_ = EAGAIN;
+ return -1;
+ }
+
+ if (type_ == SOCK_STREAM) {
+ bool was_full = (recv_buffer_size_ == server_->recv_buffer_capacity());
+ recv_buffer_size_ -= data_read;
+ if (was_full) {
+ server_->SendTcp(remote_addr_);
+ }
+ }
+
+ return data_read;
+}
+
+int VirtualSocket::SafetyBlock::RecvFrom(void* buffer,
+ size_t size,
+ SocketAddress& addr) {
+ MutexLock lock(&mutex_);
+ // If we don't have a packet, then either error or wait for one to arrive.
+ if (recv_buffer_.empty()) {
+ return -1;
+ }
+
+ // Return the packet at the front of the queue.
+ Packet& packet = *recv_buffer_.front();
+ size_t data_read = std::min(size, packet.size());
+ memcpy(buffer, packet.data(), data_read);
+ addr = packet.from();
+
+ if (data_read < packet.size()) {
+ packet.Consume(data_read);
+ } else {
+ recv_buffer_.pop_front();
+ }
+
+ // To behave like a real socket, SignalReadEvent should fire if there's still
+ // data buffered.
+ if (!recv_buffer_.empty()) {
+ PostSignalReadEvent();
+ }
+
+ return data_read;
+}
+
+int VirtualSocket::Listen(int backlog) {
+ RTC_DCHECK(SOCK_STREAM == type_);
+ RTC_DCHECK(CS_CLOSED == state_);
+ if (local_addr_.IsNil()) {
+ error_ = EINVAL;
+ return -1;
+ }
+ safety_->Listen();
+ state_ = CS_CONNECTING;
+ return 0;
+}
+
+void VirtualSocket::SafetyBlock::Listen() {
+ MutexLock lock(&mutex_);
+ RTC_DCHECK(!listen_queue_.has_value());
+ listen_queue_.emplace();
+}
+
+VirtualSocket* VirtualSocket::Accept(SocketAddress* paddr) {
+ SafetyBlock::AcceptResult result = safety_->Accept();
+ if (result.error != 0) {
+ error_ = result.error;
+ return nullptr;
+ }
+ if (paddr) {
+ *paddr = result.remote_addr;
+ }
+ return result.socket.release();
+}
+
+VirtualSocket::SafetyBlock::AcceptResult VirtualSocket::SafetyBlock::Accept() {
+ AcceptResult result;
+ MutexLock lock(&mutex_);
+ RTC_DCHECK(alive_);
+ if (!listen_queue_.has_value()) {
+ result.error = EINVAL;
+ return result;
+ }
+ while (!listen_queue_->empty()) {
+ auto socket = std::make_unique<VirtualSocket>(socket_.server_, AF_INET,
+ socket_.type_);
+
+ // Set the new local address to the same as this server socket.
+ socket->SetLocalAddress(socket_.local_addr_);
+ // Sockets made from a socket that 'was Any' need to inherit that.
+ socket->set_was_any(socket_.was_any());
+ SocketAddress remote_addr = listen_queue_->front();
+ listen_queue_->pop_front();
+ if (socket->InitiateConnect(remote_addr, false) != 0) {
+ continue;
+ }
+ socket->CompleteConnect(remote_addr);
+ result.socket = std::move(socket);
+ result.remote_addr = remote_addr;
+ return result;
+ }
+ result.error = EWOULDBLOCK;
+ return result;
+}
+
+int VirtualSocket::GetError() const {
+ return error_;
+}
+
+void VirtualSocket::SetError(int error) {
+ error_ = error;
+}
+
+Socket::ConnState VirtualSocket::GetState() const {
+ return state_;
+}
+
+int VirtualSocket::GetOption(Option opt, int* value) {
+ OptionsMap::const_iterator it = options_map_.find(opt);
+ if (it == options_map_.end()) {
+ return -1;
+ }
+ *value = it->second;
+ return 0; // 0 is success to emulate getsockopt()
+}
+
+int VirtualSocket::SetOption(Option opt, int value) {
+ options_map_[opt] = value;
+ return 0; // 0 is success to emulate setsockopt()
+}
+
+void VirtualSocket::PostPacket(TimeDelta delay,
+ std::unique_ptr<Packet> packet) {
+ rtc::scoped_refptr<SafetyBlock> safety = safety_;
+ VirtualSocket* socket = this;
+ server_->msg_queue_->PostDelayedTask(
+ [safety = std::move(safety), socket,
+ packet = std::move(packet)]() mutable {
+ if (safety->AddPacket(std::move(packet))) {
+ socket->SignalReadEvent(socket);
+ }
+ },
+ delay);
+}
+
+bool VirtualSocket::SafetyBlock::AddPacket(std::unique_ptr<Packet> packet) {
+ MutexLock lock(&mutex_);
+ if (alive_) {
+ recv_buffer_.push_back(std::move(packet));
+ }
+ return alive_;
+}
+
+void VirtualSocket::PostConnect(TimeDelta delay,
+ const SocketAddress& remote_addr) {
+ safety_->PostConnect(delay, remote_addr);
+}
+
+void VirtualSocket::SafetyBlock::PostConnect(TimeDelta delay,
+ const SocketAddress& remote_addr) {
+ rtc::scoped_refptr<SafetyBlock> safety(this);
+
+ MutexLock lock(&mutex_);
+ RTC_DCHECK(alive_);
+ // Save addresses of the pending connects to allow propertly disconnect them
+ // if socket closes before delayed task below runs.
+ // `posted_connects_` is an std::list, thus its iterators are valid while the
+ // element is in the list. It can be removed either in the `Connect` just
+ // below or by calling SetNotAlive function, thus inside `Connect` `it` should
+ // be valid when alive_ == true.
+ auto it = posted_connects_.insert(posted_connects_.end(), remote_addr);
+ auto task = [safety = std::move(safety), it] {
+ switch (safety->Connect(it)) {
+ case Signal::kNone:
+ break;
+ case Signal::kReadEvent:
+ safety->socket_.SignalReadEvent(&safety->socket_);
+ break;
+ case Signal::kConnectEvent:
+ safety->socket_.SignalConnectEvent(&safety->socket_);
+ break;
+ }
+ };
+ socket_.server_->msg_queue_->PostDelayedTask(std::move(task), delay);
+}
+
+VirtualSocket::SafetyBlock::Signal VirtualSocket::SafetyBlock::Connect(
+ VirtualSocket::SafetyBlock::PostedConnects::iterator remote_addr_it) {
+ MutexLock lock(&mutex_);
+ if (!alive_) {
+ return Signal::kNone;
+ }
+ RTC_DCHECK(!posted_connects_.empty());
+ SocketAddress remote_addr = *remote_addr_it;
+ posted_connects_.erase(remote_addr_it);
+
+ if (listen_queue_.has_value()) {
+ listen_queue_->push_back(remote_addr);
+ return Signal::kReadEvent;
+ }
+ if (socket_.type_ == SOCK_STREAM && socket_.state_ == CS_CONNECTING) {
+ socket_.CompleteConnect(remote_addr);
+ return Signal::kConnectEvent;
+ }
+ RTC_LOG(LS_VERBOSE) << "Socket at " << socket_.local_addr_.ToString()
+ << " is not listening";
+ socket_.server_->Disconnect(remote_addr);
+ return Signal::kNone;
+}
+
+bool VirtualSocket::SafetyBlock::IsAlive() {
+ MutexLock lock(&mutex_);
+ return alive_;
+}
+
+void VirtualSocket::PostDisconnect(TimeDelta delay) {
+ // Posted task may outlive this. Use different name for `this` inside the task
+ // to avoid accidental unsafe `this->safety_` instead of safe `safety`
+ VirtualSocket* socket = this;
+ rtc::scoped_refptr<SafetyBlock> safety = safety_;
+ auto task = [safety = std::move(safety), socket] {
+ if (!safety->IsAlive()) {
+ return;
+ }
+ RTC_DCHECK_EQ(socket->type_, SOCK_STREAM);
+ if (socket->state_ == CS_CLOSED) {
+ return;
+ }
+ int error_to_signal = (socket->state_ == CS_CONNECTING) ? ECONNREFUSED : 0;
+ socket->state_ = CS_CLOSED;
+ socket->remote_addr_.Clear();
+ socket->SignalCloseEvent(socket, error_to_signal);
+ };
+ server_->msg_queue_->PostDelayedTask(std::move(task), delay);
+}
+
+int VirtualSocket::InitiateConnect(const SocketAddress& addr, bool use_delay) {
+ if (!remote_addr_.IsNil()) {
+ error_ = (CS_CONNECTED == state_) ? EISCONN : EINPROGRESS;
+ return -1;
+ }
+ if (local_addr_.IsNil()) {
+ // If there's no local address set, grab a random one in the correct AF.
+ int result = 0;
+ if (addr.ipaddr().family() == AF_INET) {
+ result = Bind(SocketAddress("0.0.0.0", 0));
+ } else if (addr.ipaddr().family() == AF_INET6) {
+ result = Bind(SocketAddress("::", 0));
+ }
+ if (result != 0) {
+ return result;
+ }
+ }
+ if (type_ == SOCK_DGRAM) {
+ remote_addr_ = addr;
+ state_ = CS_CONNECTED;
+ } else {
+ int result = server_->Connect(this, addr, use_delay);
+ if (result != 0) {
+ error_ = EHOSTUNREACH;
+ return -1;
+ }
+ state_ = CS_CONNECTING;
+ }
+ return 0;
+}
+
+void VirtualSocket::CompleteConnect(const SocketAddress& addr) {
+ RTC_DCHECK(CS_CONNECTING == state_);
+ remote_addr_ = addr;
+ state_ = CS_CONNECTED;
+ server_->AddConnection(remote_addr_, local_addr_, this);
+}
+
+int VirtualSocket::SendUdp(const void* pv,
+ size_t cb,
+ const SocketAddress& addr) {
+ // If we have not been assigned a local port, then get one.
+ if (local_addr_.IsNil()) {
+ local_addr_ = server_->AssignBindAddress(
+ EmptySocketAddressWithFamily(addr.ipaddr().family()));
+ int result = server_->Bind(this, local_addr_);
+ if (result != 0) {
+ local_addr_.Clear();
+ error_ = EADDRINUSE;
+ return result;
+ }
+ }
+
+ // Send the data in a message to the appropriate socket.
+ return server_->SendUdp(this, static_cast<const char*>(pv), cb, addr);
+}
+
+int VirtualSocket::SendTcp(const void* pv, size_t cb) {
+ size_t capacity = server_->send_buffer_capacity() - send_buffer_.size();
+ if (0 == capacity) {
+ ready_to_send_ = false;
+ error_ = EWOULDBLOCK;
+ return -1;
+ }
+ size_t consumed = std::min(cb, capacity);
+ const char* cpv = static_cast<const char*>(pv);
+ send_buffer_.insert(send_buffer_.end(), cpv, cpv + consumed);
+ server_->SendTcp(this);
+ return static_cast<int>(consumed);
+}
+
+void VirtualSocket::OnSocketServerReadyToSend() {
+ if (ready_to_send_) {
+ // This socket didn't encounter EWOULDBLOCK, so there's nothing to do.
+ return;
+ }
+ if (type_ == SOCK_DGRAM) {
+ ready_to_send_ = true;
+ SignalWriteEvent(this);
+ } else {
+ RTC_DCHECK(type_ == SOCK_STREAM);
+ // This will attempt to empty the full send buffer, and will fire
+ // SignalWriteEvent if successful.
+ server_->SendTcp(this);
+ }
+}
+
+void VirtualSocket::SetToBlocked() {
+ ready_to_send_ = false;
+ error_ = EWOULDBLOCK;
+}
+
+void VirtualSocket::UpdateRecv(size_t data_size) {
+ recv_buffer_size_ += data_size;
+}
+
+void VirtualSocket::UpdateSend(size_t data_size) {
+ size_t new_buffer_size = send_buffer_.size() - data_size;
+ // Avoid undefined access beyond the last element of the vector.
+ // This only happens when new_buffer_size is 0.
+ if (data_size < send_buffer_.size()) {
+ // memmove is required for potentially overlapping source/destination.
+ memmove(&send_buffer_[0], &send_buffer_[data_size], new_buffer_size);
+ }
+ send_buffer_.resize(new_buffer_size);
+}
+
+void VirtualSocket::MaybeSignalWriteEvent(size_t capacity) {
+ if (!ready_to_send_ && (send_buffer_.size() < capacity)) {
+ ready_to_send_ = true;
+ SignalWriteEvent(this);
+ }
+}
+
+uint32_t VirtualSocket::AddPacket(int64_t cur_time, size_t packet_size) {
+ network_size_ += packet_size;
+ uint32_t send_delay =
+ server_->SendDelay(static_cast<uint32_t>(network_size_));
+
+ NetworkEntry entry;
+ entry.size = packet_size;
+ entry.done_time = cur_time + send_delay;
+ network_.push_back(entry);
+
+ return send_delay;
+}
+
+int64_t VirtualSocket::UpdateOrderedDelivery(int64_t ts) {
+ // Ensure that new packets arrive after previous ones
+ ts = std::max(ts, last_delivery_time_);
+ // A socket should not have both ordered and unordered delivery, so its last
+ // delivery time only needs to be updated when it has ordered delivery.
+ last_delivery_time_ = ts;
+ return ts;
+}
+
+size_t VirtualSocket::PurgeNetworkPackets(int64_t cur_time) {
+ while (!network_.empty() && (network_.front().done_time <= cur_time)) {
+ RTC_DCHECK(network_size_ >= network_.front().size);
+ network_size_ -= network_.front().size;
+ network_.pop_front();
+ }
+ return network_size_;
+}
+
+VirtualSocketServer::VirtualSocketServer() : VirtualSocketServer(nullptr) {}
+
+VirtualSocketServer::VirtualSocketServer(ThreadProcessingFakeClock* fake_clock)
+ : fake_clock_(fake_clock),
+ msg_queue_(nullptr),
+ stop_on_idle_(false),
+ next_ipv4_(kInitialNextIPv4),
+ next_ipv6_(kInitialNextIPv6),
+ next_port_(kFirstEphemeralPort),
+ bindings_(new AddressMap()),
+ connections_(new ConnectionMap()),
+ bandwidth_(0),
+ network_capacity_(kDefaultNetworkCapacity),
+ send_buffer_capacity_(kDefaultTcpBufferSize),
+ recv_buffer_capacity_(kDefaultTcpBufferSize),
+ delay_mean_(0),
+ delay_stddev_(0),
+ delay_samples_(NUM_SAMPLES),
+ drop_prob_(0.0) {
+ UpdateDelayDistribution();
+}
+
+VirtualSocketServer::~VirtualSocketServer() {
+ delete bindings_;
+ delete connections_;
+}
+
+IPAddress VirtualSocketServer::GetNextIP(int family) {
+ if (family == AF_INET) {
+ IPAddress next_ip(next_ipv4_);
+ next_ipv4_.s_addr = HostToNetwork32(NetworkToHost32(next_ipv4_.s_addr) + 1);
+ return next_ip;
+ } else if (family == AF_INET6) {
+ IPAddress next_ip(next_ipv6_);
+ uint32_t* as_ints = reinterpret_cast<uint32_t*>(&next_ipv6_.s6_addr);
+ as_ints[3] += 1;
+ return next_ip;
+ }
+ return IPAddress();
+}
+
+uint16_t VirtualSocketServer::GetNextPort() {
+ uint16_t port = next_port_;
+ if (next_port_ < kLastEphemeralPort) {
+ ++next_port_;
+ } else {
+ next_port_ = kFirstEphemeralPort;
+ }
+ return port;
+}
+
+void VirtualSocketServer::SetSendingBlocked(bool blocked) {
+ {
+ webrtc::MutexLock lock(&mutex_);
+ if (blocked == sending_blocked_) {
+ // Unchanged; nothing to do.
+ return;
+ }
+ sending_blocked_ = blocked;
+ }
+ if (!blocked) {
+ // Sending was blocked, but is now unblocked. This signal gives sockets a
+ // chance to fire SignalWriteEvent, and for TCP, send buffered data.
+ SignalReadyToSend();
+ }
+}
+
+VirtualSocket* VirtualSocketServer::CreateSocket(int family, int type) {
+ return new VirtualSocket(this, family, type);
+}
+
+void VirtualSocketServer::SetMessageQueue(Thread* msg_queue) {
+ msg_queue_ = msg_queue;
+}
+
+bool VirtualSocketServer::Wait(webrtc::TimeDelta max_wait_duration,
+ bool process_io) {
+ RTC_DCHECK_RUN_ON(msg_queue_);
+ if (stop_on_idle_ && Thread::Current()->empty()) {
+ return false;
+ }
+ // Note: we don't need to do anything with `process_io` since we don't have
+ // any real I/O. Received packets come in the form of queued messages, so
+ // Thread will ensure WakeUp is called if another thread sends a
+ // packet.
+ wakeup_.Wait(max_wait_duration);
+ return true;
+}
+
+void VirtualSocketServer::WakeUp() {
+ wakeup_.Set();
+}
+
+void VirtualSocketServer::SetAlternativeLocalAddress(
+ const rtc::IPAddress& address,
+ const rtc::IPAddress& alternative) {
+ alternative_address_mapping_[address] = alternative;
+}
+
+bool VirtualSocketServer::ProcessMessagesUntilIdle() {
+ RTC_DCHECK_RUN_ON(msg_queue_);
+ stop_on_idle_ = true;
+ while (!msg_queue_->empty()) {
+ if (fake_clock_) {
+ // If using a fake clock, advance it in millisecond increments until the
+ // queue is empty.
+ fake_clock_->AdvanceTime(webrtc::TimeDelta::Millis(1));
+ } else {
+ // Otherwise, run a normal message loop.
+ msg_queue_->ProcessMessages(Thread::kForever);
+ }
+ }
+ stop_on_idle_ = false;
+ return !msg_queue_->IsQuitting();
+}
+
+void VirtualSocketServer::SetNextPortForTesting(uint16_t port) {
+ next_port_ = port;
+}
+
+bool VirtualSocketServer::CloseTcpConnections(
+ const SocketAddress& addr_local,
+ const SocketAddress& addr_remote) {
+ VirtualSocket* socket = LookupConnection(addr_local, addr_remote);
+ if (!socket) {
+ return false;
+ }
+ // Signal the close event on the local connection first.
+ socket->SignalCloseEvent(socket, 0);
+
+ // Trigger the remote connection's close event.
+ socket->Close();
+
+ return true;
+}
+
+int VirtualSocketServer::Bind(VirtualSocket* socket,
+ const SocketAddress& addr) {
+ RTC_DCHECK(nullptr != socket);
+ // Address must be completely specified at this point
+ RTC_DCHECK(!IPIsUnspec(addr.ipaddr()));
+ RTC_DCHECK(addr.port() != 0);
+
+ // Normalize the address (turns v6-mapped addresses into v4-addresses).
+ SocketAddress normalized(addr.ipaddr().Normalized(), addr.port());
+
+ AddressMap::value_type entry(normalized, socket);
+ return bindings_->insert(entry).second ? 0 : -1;
+}
+
+SocketAddress VirtualSocketServer::AssignBindAddress(
+ const SocketAddress& app_addr) {
+ RTC_DCHECK(!IPIsUnspec(app_addr.ipaddr()));
+
+ // Normalize the IP.
+ SocketAddress addr;
+ addr.SetIP(app_addr.ipaddr().Normalized());
+
+ // If the IP appears in `alternative_address_mapping_`, meaning the test has
+ // configured sockets bound to this IP to actually use another IP, replace
+ // the IP here.
+ auto alternative = alternative_address_mapping_.find(addr.ipaddr());
+ if (alternative != alternative_address_mapping_.end()) {
+ addr.SetIP(alternative->second);
+ }
+
+ if (app_addr.port() != 0) {
+ addr.SetPort(app_addr.port());
+ } else {
+ // Assign a port.
+ for (int i = 0; i < kEphemeralPortCount; ++i) {
+ addr.SetPort(GetNextPort());
+ if (bindings_->find(addr) == bindings_->end()) {
+ break;
+ }
+ }
+ }
+
+ return addr;
+}
+
+VirtualSocket* VirtualSocketServer::LookupBinding(const SocketAddress& addr) {
+ SocketAddress normalized(addr.ipaddr().Normalized(), addr.port());
+ AddressMap::iterator it = bindings_->find(normalized);
+ if (it != bindings_->end()) {
+ return it->second;
+ }
+
+ IPAddress default_ip = GetDefaultSourceAddress(addr.ipaddr().family());
+ if (!IPIsUnspec(default_ip) && addr.ipaddr() == default_ip) {
+ // If we can't find a binding for the packet which is sent to the interface
+ // corresponding to the default route, it should match a binding with the
+ // correct port to the any address.
+ SocketAddress sock_addr =
+ EmptySocketAddressWithFamily(addr.ipaddr().family());
+ sock_addr.SetPort(addr.port());
+ return LookupBinding(sock_addr);
+ }
+
+ return nullptr;
+}
+
+int VirtualSocketServer::Unbind(const SocketAddress& addr,
+ VirtualSocket* socket) {
+ SocketAddress normalized(addr.ipaddr().Normalized(), addr.port());
+ RTC_DCHECK((*bindings_)[normalized] == socket);
+ bindings_->erase(bindings_->find(normalized));
+ return 0;
+}
+
+void VirtualSocketServer::AddConnection(const SocketAddress& local,
+ const SocketAddress& remote,
+ VirtualSocket* remote_socket) {
+ // Add this socket pair to our routing table. This will allow
+ // multiple clients to connect to the same server address.
+ SocketAddress local_normalized(local.ipaddr().Normalized(), local.port());
+ SocketAddress remote_normalized(remote.ipaddr().Normalized(), remote.port());
+ SocketAddressPair address_pair(local_normalized, remote_normalized);
+ connections_->insert(std::pair<SocketAddressPair, VirtualSocket*>(
+ address_pair, remote_socket));
+}
+
+VirtualSocket* VirtualSocketServer::LookupConnection(
+ const SocketAddress& local,
+ const SocketAddress& remote) {
+ SocketAddress local_normalized(local.ipaddr().Normalized(), local.port());
+ SocketAddress remote_normalized(remote.ipaddr().Normalized(), remote.port());
+ SocketAddressPair address_pair(local_normalized, remote_normalized);
+ ConnectionMap::iterator it = connections_->find(address_pair);
+ return (connections_->end() != it) ? it->second : nullptr;
+}
+
+void VirtualSocketServer::RemoveConnection(const SocketAddress& local,
+ const SocketAddress& remote) {
+ SocketAddress local_normalized(local.ipaddr().Normalized(), local.port());
+ SocketAddress remote_normalized(remote.ipaddr().Normalized(), remote.port());
+ SocketAddressPair address_pair(local_normalized, remote_normalized);
+ connections_->erase(address_pair);
+}
+
+static double Random() {
+ return static_cast<double>(rand()) / RAND_MAX;
+}
+
+int VirtualSocketServer::Connect(VirtualSocket* socket,
+ const SocketAddress& remote_addr,
+ bool use_delay) {
+ RTC_DCHECK(msg_queue_);
+
+ TimeDelta delay = TimeDelta::Millis(use_delay ? GetTransitDelay(socket) : 0);
+ VirtualSocket* remote = LookupBinding(remote_addr);
+ if (!CanInteractWith(socket, remote)) {
+ RTC_LOG(LS_INFO) << "Address family mismatch between "
+ << socket->GetLocalAddress().ToString() << " and "
+ << remote_addr.ToString();
+ return -1;
+ }
+ if (remote != nullptr) {
+ remote->PostConnect(delay, socket->GetLocalAddress());
+ } else {
+ RTC_LOG(LS_INFO) << "No one listening at " << remote_addr.ToString();
+ socket->PostDisconnect(delay);
+ }
+ return 0;
+}
+
+bool VirtualSocketServer::Disconnect(VirtualSocket* socket) {
+ if (!socket || !msg_queue_)
+ return false;
+
+ // If we simulate packets being delayed, we should simulate the
+ // equivalent of a FIN being delayed as well.
+ socket->PostDisconnect(TimeDelta::Millis(GetTransitDelay(socket)));
+ return true;
+}
+
+bool VirtualSocketServer::Disconnect(const SocketAddress& addr) {
+ return Disconnect(LookupBinding(addr));
+}
+
+bool VirtualSocketServer::Disconnect(const SocketAddress& local_addr,
+ const SocketAddress& remote_addr) {
+ // Disconnect remote socket, check if it is a child of a server socket.
+ VirtualSocket* socket = LookupConnection(local_addr, remote_addr);
+ if (!socket) {
+ // Not a server socket child, then see if it is bound.
+ // TODO(tbd): If this is indeed a server socket that has no
+ // children this will cause the server socket to be
+ // closed. This might lead to unexpected results, how to fix this?
+ socket = LookupBinding(remote_addr);
+ }
+ Disconnect(socket);
+
+ // Remove mapping for both directions.
+ RemoveConnection(remote_addr, local_addr);
+ RemoveConnection(local_addr, remote_addr);
+ return socket != nullptr;
+}
+
+int VirtualSocketServer::SendUdp(VirtualSocket* socket,
+ const char* data,
+ size_t data_size,
+ const SocketAddress& remote_addr) {
+ {
+ webrtc::MutexLock lock(&mutex_);
+ ++sent_packets_;
+ if (sending_blocked_) {
+ socket->SetToBlocked();
+ return -1;
+ }
+
+ // See if we want to drop this packet.
+ if (data_size > max_udp_payload_) {
+ RTC_LOG(LS_VERBOSE) << "Dropping too large UDP payload of size "
+ << data_size << ", UDP payload limit is "
+ << max_udp_payload_;
+ // Return as if send was successful; packet disappears.
+ return data_size;
+ }
+
+ if (Random() < drop_prob_) {
+ RTC_LOG(LS_VERBOSE) << "Dropping packet: bad luck";
+ return static_cast<int>(data_size);
+ }
+ }
+
+ VirtualSocket* recipient = LookupBinding(remote_addr);
+ if (!recipient) {
+ // Make a fake recipient for address family checking.
+ std::unique_ptr<VirtualSocket> dummy_socket(
+ CreateSocket(AF_INET, SOCK_DGRAM));
+ dummy_socket->SetLocalAddress(remote_addr);
+ if (!CanInteractWith(socket, dummy_socket.get())) {
+ RTC_LOG(LS_VERBOSE) << "Incompatible address families: "
+ << socket->GetLocalAddress().ToString() << " and "
+ << remote_addr.ToString();
+ return -1;
+ }
+ RTC_LOG(LS_VERBOSE) << "No one listening at " << remote_addr.ToString();
+ return static_cast<int>(data_size);
+ }
+
+ if (!CanInteractWith(socket, recipient)) {
+ RTC_LOG(LS_VERBOSE) << "Incompatible address families: "
+ << socket->GetLocalAddress().ToString() << " and "
+ << remote_addr.ToString();
+ return -1;
+ }
+
+ {
+ int64_t cur_time = TimeMillis();
+ size_t network_size = socket->PurgeNetworkPackets(cur_time);
+
+ // Determine whether we have enough bandwidth to accept this packet. To do
+ // this, we need to update the send queue. Once we know it's current size,
+ // we know whether we can fit this packet.
+ //
+ // NOTE: There are better algorithms for maintaining such a queue (such as
+ // "Derivative Random Drop"); however, this algorithm is a more accurate
+ // simulation of what a normal network would do.
+ {
+ webrtc::MutexLock lock(&mutex_);
+ size_t packet_size = data_size + UDP_HEADER_SIZE;
+ if (network_size + packet_size > network_capacity_) {
+ RTC_LOG(LS_VERBOSE) << "Dropping packet: network capacity exceeded";
+ return static_cast<int>(data_size);
+ }
+ }
+
+ AddPacketToNetwork(socket, recipient, cur_time, data, data_size,
+ UDP_HEADER_SIZE, false);
+
+ return static_cast<int>(data_size);
+ }
+}
+
+void VirtualSocketServer::SendTcp(VirtualSocket* socket) {
+ {
+ webrtc::MutexLock lock(&mutex_);
+ ++sent_packets_;
+ if (sending_blocked_) {
+ // Eventually the socket's buffer will fill and VirtualSocket::SendTcp
+ // will set EWOULDBLOCK.
+ return;
+ }
+ }
+
+ // TCP can't send more data than will fill up the receiver's buffer.
+ // We track the data that is in the buffer plus data in flight using the
+ // recipient's recv_buffer_size_. Anything beyond that must be stored in the
+ // sender's buffer. We will trigger the buffered data to be sent when data
+ // is read from the recv_buffer.
+
+ // Lookup the local/remote pair in the connections table.
+ VirtualSocket* recipient =
+ LookupConnection(socket->GetLocalAddress(), socket->GetRemoteAddress());
+ if (!recipient) {
+ RTC_LOG(LS_VERBOSE) << "Sending data to no one.";
+ return;
+ }
+
+ int64_t cur_time = TimeMillis();
+ socket->PurgeNetworkPackets(cur_time);
+
+ while (true) {
+ size_t available = recv_buffer_capacity() - recipient->recv_buffer_size();
+ size_t max_data_size =
+ std::min<size_t>(available, TCP_MSS - TCP_HEADER_SIZE);
+ size_t data_size = std::min(socket->send_buffer_size(), max_data_size);
+ if (0 == data_size)
+ break;
+
+ AddPacketToNetwork(socket, recipient, cur_time, socket->send_buffer_data(),
+ data_size, TCP_HEADER_SIZE, true);
+ recipient->UpdateRecv(data_size);
+ socket->UpdateSend(data_size);
+ }
+
+ socket->MaybeSignalWriteEvent(send_buffer_capacity());
+}
+
+void VirtualSocketServer::SendTcp(const SocketAddress& addr) {
+ VirtualSocket* sender = LookupBinding(addr);
+ RTC_DCHECK(nullptr != sender);
+ SendTcp(sender);
+}
+
+void VirtualSocketServer::AddPacketToNetwork(VirtualSocket* sender,
+ VirtualSocket* recipient,
+ int64_t cur_time,
+ const char* data,
+ size_t data_size,
+ size_t header_size,
+ bool ordered) {
+ RTC_DCHECK(msg_queue_);
+ uint32_t send_delay = sender->AddPacket(cur_time, data_size + header_size);
+
+ // Find the delay for crossing the many virtual hops of the network.
+ uint32_t transit_delay = GetTransitDelay(sender);
+
+ // When the incoming packet is from a binding of the any address, translate it
+ // to the default route here such that the recipient will see the default
+ // route.
+ SocketAddress sender_addr = sender->GetLocalAddress();
+ IPAddress default_ip = GetDefaultSourceAddress(sender_addr.ipaddr().family());
+ if (sender_addr.IsAnyIP() && !IPIsUnspec(default_ip)) {
+ sender_addr.SetIP(default_ip);
+ }
+
+ int64_t ts = cur_time + send_delay + transit_delay;
+ if (ordered) {
+ ts = sender->UpdateOrderedDelivery(ts);
+ }
+ recipient->PostPacket(TimeDelta::Millis(ts - cur_time),
+ std::make_unique<Packet>(data, data_size, sender_addr));
+}
+
+uint32_t VirtualSocketServer::SendDelay(uint32_t size) {
+ webrtc::MutexLock lock(&mutex_);
+ if (bandwidth_ == 0)
+ return 0;
+ else
+ return 1000 * size / bandwidth_;
+}
+
+#if 0
+void PrintFunction(std::vector<std::pair<double, double> >* f) {
+ return;
+ double sum = 0;
+ for (uint32_t i = 0; i < f->size(); ++i) {
+ std::cout << (*f)[i].first << '\t' << (*f)[i].second << std::endl;
+ sum += (*f)[i].second;
+ }
+ if (!f->empty()) {
+ const double mean = sum / f->size();
+ double sum_sq_dev = 0;
+ for (uint32_t i = 0; i < f->size(); ++i) {
+ double dev = (*f)[i].second - mean;
+ sum_sq_dev += dev * dev;
+ }
+ std::cout << "Mean = " << mean << " StdDev = "
+ << sqrt(sum_sq_dev / f->size()) << std::endl;
+ }
+}
+#endif // <unused>
+
+void VirtualSocketServer::UpdateDelayDistribution() {
+ webrtc::MutexLock lock(&mutex_);
+ delay_dist_ = CreateDistribution(delay_mean_, delay_stddev_, delay_samples_);
+}
+
+static double PI = 4 * atan(1.0);
+
+static double Normal(double x, double mean, double stddev) {
+ double a = (x - mean) * (x - mean) / (2 * stddev * stddev);
+ return exp(-a) / (stddev * sqrt(2 * PI));
+}
+
+#if 0 // static unused gives a warning
+static double Pareto(double x, double min, double k) {
+ if (x < min)
+ return 0;
+ else
+ return k * std::pow(min, k) / std::pow(x, k+1);
+}
+#endif
+
+std::unique_ptr<VirtualSocketServer::Function>
+VirtualSocketServer::CreateDistribution(uint32_t mean,
+ uint32_t stddev,
+ uint32_t samples) {
+ auto f = std::make_unique<Function>();
+
+ if (0 == stddev) {
+ f->push_back(Point(mean, 1.0));
+ } else {
+ double start = 0;
+ if (mean >= 4 * static_cast<double>(stddev))
+ start = mean - 4 * static_cast<double>(stddev);
+ double end = mean + 4 * static_cast<double>(stddev);
+
+ for (uint32_t i = 0; i < samples; i++) {
+ double x = start + (end - start) * i / (samples - 1);
+ double y = Normal(x, mean, stddev);
+ f->push_back(Point(x, y));
+ }
+ }
+ return Resample(Invert(Accumulate(std::move(f))), 0, 1, samples);
+}
+
+uint32_t VirtualSocketServer::GetTransitDelay(Socket* socket) {
+ // Use the delay based on the address if it is set.
+ auto iter = delay_by_ip_.find(socket->GetLocalAddress().ipaddr());
+ if (iter != delay_by_ip_.end()) {
+ return static_cast<uint32_t>(iter->second);
+ }
+ // Otherwise, use the delay from the distribution distribution.
+ size_t index = rand() % delay_dist_->size();
+ double delay = (*delay_dist_)[index].second;
+ // RTC_LOG_F(LS_INFO) << "random[" << index << "] = " << delay;
+ return static_cast<uint32_t>(delay);
+}
+
+struct FunctionDomainCmp {
+ bool operator()(const VirtualSocketServer::Point& p1,
+ const VirtualSocketServer::Point& p2) {
+ return p1.first < p2.first;
+ }
+ bool operator()(double v1, const VirtualSocketServer::Point& p2) {
+ return v1 < p2.first;
+ }
+ bool operator()(const VirtualSocketServer::Point& p1, double v2) {
+ return p1.first < v2;
+ }
+};
+
+std::unique_ptr<VirtualSocketServer::Function> VirtualSocketServer::Accumulate(
+ std::unique_ptr<Function> f) {
+ RTC_DCHECK(f->size() >= 1);
+ double v = 0;
+ for (Function::size_type i = 0; i < f->size() - 1; ++i) {
+ double dx = (*f)[i + 1].first - (*f)[i].first;
+ double avgy = ((*f)[i + 1].second + (*f)[i].second) / 2;
+ (*f)[i].second = v;
+ v = v + dx * avgy;
+ }
+ (*f)[f->size() - 1].second = v;
+ return f;
+}
+
+std::unique_ptr<VirtualSocketServer::Function> VirtualSocketServer::Invert(
+ std::unique_ptr<Function> f) {
+ for (Function::size_type i = 0; i < f->size(); ++i)
+ std::swap((*f)[i].first, (*f)[i].second);
+
+ absl::c_sort(*f, FunctionDomainCmp());
+ return f;
+}
+
+std::unique_ptr<VirtualSocketServer::Function> VirtualSocketServer::Resample(
+ std::unique_ptr<Function> f,
+ double x1,
+ double x2,
+ uint32_t samples) {
+ auto g = std::make_unique<Function>();
+
+ for (size_t i = 0; i < samples; i++) {
+ double x = x1 + (x2 - x1) * i / (samples - 1);
+ double y = Evaluate(f.get(), x);
+ g->push_back(Point(x, y));
+ }
+
+ return g;
+}
+
+double VirtualSocketServer::Evaluate(const Function* f, double x) {
+ Function::const_iterator iter =
+ absl::c_lower_bound(*f, x, FunctionDomainCmp());
+ if (iter == f->begin()) {
+ return (*f)[0].second;
+ } else if (iter == f->end()) {
+ RTC_DCHECK(f->size() >= 1);
+ return (*f)[f->size() - 1].second;
+ } else if (iter->first == x) {
+ return iter->second;
+ } else {
+ double x1 = (iter - 1)->first;
+ double y1 = (iter - 1)->second;
+ double x2 = iter->first;
+ double y2 = iter->second;
+ return y1 + (y2 - y1) * (x - x1) / (x2 - x1);
+ }
+}
+
+bool VirtualSocketServer::CanInteractWith(VirtualSocket* local,
+ VirtualSocket* remote) {
+ if (!local || !remote) {
+ return false;
+ }
+ IPAddress local_ip = local->GetLocalAddress().ipaddr();
+ IPAddress remote_ip = remote->GetLocalAddress().ipaddr();
+ IPAddress local_normalized = local_ip.Normalized();
+ IPAddress remote_normalized = remote_ip.Normalized();
+ // Check if the addresses are the same family after Normalization (turns
+ // mapped IPv6 address into IPv4 addresses).
+ // This will stop unmapped V6 addresses from talking to mapped V6 addresses.
+ if (local_normalized.family() == remote_normalized.family()) {
+ return true;
+ }
+
+ // If ip1 is IPv4 and ip2 is :: and ip2 is not IPV6_V6ONLY.
+ int remote_v6_only = 0;
+ remote->GetOption(Socket::OPT_IPV6_V6ONLY, &remote_v6_only);
+ if (local_ip.family() == AF_INET && !remote_v6_only && IPIsAny(remote_ip)) {
+ return true;
+ }
+ // Same check, backwards.
+ int local_v6_only = 0;
+ local->GetOption(Socket::OPT_IPV6_V6ONLY, &local_v6_only);
+ if (remote_ip.family() == AF_INET && !local_v6_only && IPIsAny(local_ip)) {
+ return true;
+ }
+
+ // Check to see if either socket was explicitly bound to IPv6-any.
+ // These sockets can talk with anyone.
+ if (local_ip.family() == AF_INET6 && local->was_any()) {
+ return true;
+ }
+ if (remote_ip.family() == AF_INET6 && remote->was_any()) {
+ return true;
+ }
+
+ return false;
+}
+
+IPAddress VirtualSocketServer::GetDefaultSourceAddress(int family) {
+ if (family == AF_INET) {
+ return default_source_address_v4_;
+ }
+ if (family == AF_INET6) {
+ return default_source_address_v6_;
+ }
+ return IPAddress();
+}
+void VirtualSocketServer::SetDefaultSourceAddress(const IPAddress& from_addr) {
+ RTC_DCHECK(!IPIsAny(from_addr));
+ if (from_addr.family() == AF_INET) {
+ default_source_address_v4_ = from_addr;
+ } else if (from_addr.family() == AF_INET6) {
+ default_source_address_v6_ = from_addr;
+ }
+}
+
+void VirtualSocketServer::set_bandwidth(uint32_t bandwidth) {
+ webrtc::MutexLock lock(&mutex_);
+ bandwidth_ = bandwidth;
+}
+void VirtualSocketServer::set_network_capacity(uint32_t capacity) {
+ webrtc::MutexLock lock(&mutex_);
+ network_capacity_ = capacity;
+}
+
+uint32_t VirtualSocketServer::send_buffer_capacity() const {
+ webrtc::MutexLock lock(&mutex_);
+ return send_buffer_capacity_;
+}
+void VirtualSocketServer::set_send_buffer_capacity(uint32_t capacity) {
+ webrtc::MutexLock lock(&mutex_);
+ send_buffer_capacity_ = capacity;
+}
+
+uint32_t VirtualSocketServer::recv_buffer_capacity() const {
+ webrtc::MutexLock lock(&mutex_);
+ return recv_buffer_capacity_;
+}
+void VirtualSocketServer::set_recv_buffer_capacity(uint32_t capacity) {
+ webrtc::MutexLock lock(&mutex_);
+ recv_buffer_capacity_ = capacity;
+}
+
+void VirtualSocketServer::set_delay_mean(uint32_t delay_mean) {
+ webrtc::MutexLock lock(&mutex_);
+ delay_mean_ = delay_mean;
+}
+void VirtualSocketServer::set_delay_stddev(uint32_t delay_stddev) {
+ webrtc::MutexLock lock(&mutex_);
+ delay_stddev_ = delay_stddev;
+}
+void VirtualSocketServer::set_delay_samples(uint32_t delay_samples) {
+ webrtc::MutexLock lock(&mutex_);
+ delay_samples_ = delay_samples;
+}
+
+void VirtualSocketServer::set_drop_probability(double drop_prob) {
+ RTC_DCHECK_GE(drop_prob, 0.0);
+ RTC_DCHECK_LE(drop_prob, 1.0);
+
+ webrtc::MutexLock lock(&mutex_);
+ drop_prob_ = drop_prob;
+}
+
+void VirtualSocketServer::set_max_udp_payload(size_t payload_size) {
+ webrtc::MutexLock lock(&mutex_);
+ max_udp_payload_ = payload_size;
+}
+
+uint32_t VirtualSocketServer::sent_packets() const {
+ webrtc::MutexLock lock(&mutex_);
+ return sent_packets_;
+}
+
+} // namespace rtc